My Experience

Nasr City, Cairo

• Extended the finance module by adding e-invoicing in multiple regions via integrations with ETA (Egypt) and ZATCA (Saudi Arabia).
• Enhanced CI/CD workflows via GitHub Actions, reducing build time by 41%.
• Containerized the ASP.NET, Angular, and SQL Server stack as a multi-container Docker application.
• Architected a multi-agent AI platform (FastAPI, LangGraph, LangChain) for enterprise ERP with LLM-powered SQL analysis, document understanding, and sandboxed code execution.
• Implemented RAG pipelines using FAISS and ChromaDB for domain-specific retrieval and contextual memory augmentation.
• Integrated Langfuse observability for LLM tracing and monitoring; optimized LangGraph state storage reducing size by 70%.

• Developed a machine learning pipeline for predicting the QOS (quality of service) of a 5G mobile network using environmental variables (temperature, longitude, latitude) and RF measurements (RSRP, SNR, bandwidth).
• Trained the pipeline on the Berlin V2X dataset achieving 7.97 mbit/s RMSE in predicting the QOS of the test split.
• The pipeline used CatBoostRegressor and incorporated feature engineering (one hot encoding, lag features) and hyperparameter tuning.

• Used ROS (Robot Operating System), the moveit library, and OpenCV to control a robotic arm via Arduino, utilizing camera feed for input.